Author Correction: Engineering novel CRISPRi repressors for highly efficient mammalian gene regulation

Correction: Genome Biology 26, 164 (2025)

https://doi.org/10.1186/s13059-025-03640-4

Following publication of the original article, “Engineering novel CRISPRi repressors for highly efficient mammalian gene regulation” [1], authors have added an additional citation to a previous work [2] that demonstrates that dCas9-Zim3-MECP2 leads to efficient and long-term epigenetic silencing.

Additionally, the authors have made the following changes to the text to reflect the contributions of the previous work.

Figure changes:

(1) The bar colors in Fig. 1D and 1F have been fixed to show that dCas9-ZIM3-MeCP2 is a previously published CRISPRi system.

Original:

Corrected:

(2) Supplementary Figure S3 and its caption have been updated to change the color of the previously characterized dCas9-ZIM3-MeCP2 to be gray such that it is different than the three novel bipartite variants to prevent possible confusion about its novelty.

Original:

Supplementary Figure S3. Gene silencing achieved by improved bipartite repressor domains. (A) Comparison of top-performing prior CRISPRi platforms and novel bipartite repressor variants in HEK293T cells. Indicated dCas9-repressor fusions and eGFP reporter construct (with alternative sgRNA −313(T) targeting the SV40 promoter) were co-transfected and samples were assayed 72 h later using flow cytometry. Wild type (WT) cells indicate level of complete eGFP silencing. (B) Histograms showing the distribution of eGFP fluorescence for current “gold standards” and one novel variant, KOX1(KRAB)-MeCP2(t) compared to cells expressing only dCas9 (solid line, tan population) and WT cells (dashed line, white population). Percent silencing for each dCas9-repressor fusion was quantified by assessing the number of cells in each population overlapping the WT cells population.

Corrected:

Supplementary Figure S3. Gene silencing achieved by bipartite repressor domains. (A) Comparison of top-performing prior CRISPRi platforms and top-performing bipartite repressor variants in HEK293T cells. Indicated dCas9-repressor fusions and eGFP reporter construct (with alternative sgRNA −313(T) targeting the SV40 promoter) were co-transfected and samples were assayed 72 h later using flow cytometry. Wild type (WT) cells indicate level of complete eGFP silencing. (B) Histograms showing the distribution of eGFP fluorescence for current “gold standards” and one novel variant, KOX1(KRAB)-MeCP2(t) compared to cells expressing only dCas9 (solid line, tan population) and WT cells (dashed line, white population). Percent silencing for each dCas9-repressor fusion was quantified by assessing the number of cells in each population overlapping the WT cells population.

(3) Supplementary Figure S4 has been updated to change the color of the previously characterized dCas9-ZIM3-MeCP2 to light gray to note that it was not used for more experiments that are presented in the main text figures.

Original:

Corrected:

Main Text Changes:

Change #1:

Original: Four novel repressor combinations (dCas9-KRBOX1(KRAB)-MAX, dCas9-ZIM3(KRAB)-MAX, dCas9-ZIM3(KRAB)-MeCP2, and dCas9-KOX1(KRAB)-MeCP2(t)) significantly improved knockdown (~ 20–30% better, p* < 0.05) and silencing percentage compared to dCas9-ZIM3(KRAB), the top performing CRISPRi system created and characterized previously, in our tests in HEK293T cells (Fig. 1D, Additional File 2: Supplementary Figure S3).

Edited: Three novel repressor combinations (dCas9-KRBOX1(KRAB)-MAX, dCas9-ZIM3(KRAB)-MAX, and dCas9-KOX1(KRAB)-MeCP2(t)), and one previously tested variant also found in our screen (dCas9-ZIM3(KRAB)-MeCP2) [2], significantly improved gene knockdown (~ 20–30% better, p* < 0.05) compared to dCas9-ZIM3(KRAB) in our tests in HEK293T cells (Fig. 1D, Additional File 2: Supplementary Figure S3).

Change #2:

Original: Also, the improved gene knockdown ability of these novel bipartite dCas9-repressor fusions did not correlate with their expression levels, as several had lower expression than dCas9-ZIM3(KRAB) and there was no correlation between dCas9-repressor expression and eGFP knockdown in the initial library screen (Additional File 2: Supplementary Figure S4 A).

Edited: Also, the improved gene knockdown ability of these four bipartite dCas9-repressor fusions did not correlate with their expression levels, as several had lower expression than dCas9-ZIM3(KRAB) and there was no correlation between dCas9-repressor expression and eGFP knockdown in the initial library screen (Additional File 2: Supplementary Figure S4A).

Change #3:

Original: As our library results suggested that (1) further addition of domains to our combinatorial repressor constructs may fail to increase their potency and (2) the five novel repressor fusions we have highlighted (4 bipartite and 1 tripartite) appeared to outperform prior gold standard CRISPRi repressor proteins, we next sought to evaluate these novel fusions more fully.

Edited: As our library results suggested that (1) further addition of domains to our combinatorial repressor constructs may fail to increase their potency and (2) the highly potent repressor fusions we have highlighted (4 bipartite and 1 tripartite) appeared to outperform gold standard CRISPRi repressor proteins, we next sought to evaluate a subset these fusions more fully.

Change #4:

Original: The gold standard repressor domains we utilized for comparative analyses to our novel variants had identical protein sequences to those developed previously [32, 35].

Edited: The gold standard repressor domains we utilized for comparative analyses to our variants had identical protein sequences to those developed previously [32, 35].

Change #5:

Original: It also suggests that our novel variants’ high activity was unlikely to be caused by the dual-targeting reporter sgRNA used in our preliminary screens.

Edited: It also suggests that our variants’ high activity was unlikely to be caused by the dual-targeting reporter sgRNA used in our preliminary screens.

These changes do not affect the main results and conclusions of the paper.

The HTML and PDF versions of the original article [1] have been updated.